| dc.description.abstract | The epitaxial growth of perovskite-structure oxide thin films on silicon (Si) substrates presents an avenue for integrating diverse electronic, optoelectronic, and acoustic functionalities into the well-established, cost-effective Si-based integrated circuit technology. BaTiO3 (BTO) is extensively studied due to its notable dielectric, ferroelectric, piezoelectric, and non-linear optic characteristics, with potential applications in integrated devices and optical modulators. However, the direct integration of BTO onto Si encounters challenges stemming from both the oxidation of the silicon surface, the chemical reactivity at the BTO/Si interface, and the substantial lattice mismatch between BTO and Si (4.43%). Various buffer layer combinations are necessary to address these challenges, including SRO/MgO/TiN//Si, LNO/STO//Si, and LNO/CeO2/YSZ/Si, among others. The growth of epitaxial STO templates on Si demands complex synthesis methods such as molecular beam epitaxy (MBE) and ultra-high vacuum (UHV) conditions. However, growing BTO on lattice-matched single crystalline oxide substrates like STO is relatively straightforward. However, STO substrates are costly and not CMOS-compatible. Here, we ask whether it is possible to epitaxially grow BTO on a lattice-matched single crystalline STO substrate, followed by its transfer onto Si. Conventional methods for epitaxial transfer typically involve mechanical exfoliation, chemical etching, or optically induced separation between the epilayer and the substrate. However, these methods are often plagued by rough interfaces, challenges in achieving selective etching between layers, and a lack of precise control over spalling depth.
This thesis discusses alternative techniques for epitaxial transfer of functional oxides onto silicon and flexible substrates. This study examines two primary approaches for transferring BTO from STO to silicon: one involving a water-soluble sacrificial buffer layer (such as SrVO3 (SVO) or Sr3Al2O3 (SAO)), and the other employing a 2D material at the interface between the film and the substrate, known as remote epitaxy.
A buffer layer-free approach to integrate BTO on Si is presented. BTO was grown on a single crystalline STO substrate via a water-soluble SAO sacrificial layer, and the BTO layer was subsequently transferred onto Si using a simple technique. The dissolution and transfer process takes less than an hour in total. The ferroelectricity (evidenced by polarization-electric field loops) and piezoelectricity (displacement versus voltage butterfly curve obtained from a Laser Doppler Vibrometer) of the transferred BTO on Si were demonstrated through comprehensive measurements. Thicker BTO disintegrates while transferring; thus, the concept of a templated substrate was introduced, where a 5mm x 5 mm thin BTO membrane was transferred onto Si, termed as a templated Si substrate, and showcased the epitaxial growth of BTO on this templated substrate. Utilizing high-resolution X-ray diffraction, atomic force microscopy, and scanning transmission electron microscopy, the single-crystalline nature of the BTO film on Si was exhibited, and polarization distribution was obtained.
One challenge associated with the SAO layer is its susceptibility to atmospheric moisture, rendering non-vacuum-based deposition of complex oxides on top of it unfeasible. A stable (less reactive) sacrificial layer SVO was introduced to improve the versatility of layer transfer techniques. Towards the second chapter of this thesis, a combination of chemical vapor deposited (CVD) graphene as a 2D material at the interface and pulsed laser-deposited (PLD) water-soluble SrVO3 (SVO) as a sacrificial buffer layer was utilized. Thereafter, the well-known enhancement of liquid diffusivities by monolayer graphene enhances the dissolution rate of SVO over ten times without compromising its atmospheric stability. The versatility of the hybrid graphene-SVO//STO template was demonstrated by growing ferroelectric BTO via PLD and Pb(Zr, Ti)O3 (PZT) via Chemical Solution Deposition (CSD) technique and transferring them onto the target substrates and establishing their ferroelectric properties.
The phenomenon of BaO leaching in water is documented in bulk BTO literature, albeit with very low rates. Towards the final chapter of this thesis, a non-aqueous-based transfer method is introduced. Remote epitaxy has garnered considerable attention as a promising method recently, which facilitates the growth of thin films that replicate the crystallographic characteristics of the substrate by utilizing two-dimensional material interlayers. The resulting film can be exfoliated to form a freestanding membrane, although it is often challenging since two-dimensional materials are susceptible to damage under harsh epitaxy conditions. In contrast to nitrides remote epitaxy, the popularity of oxide remote epitaxy is comparatively lower primarily because the conventional aggressive oxidizing conditions employed for the growth of epitaxial oxides pose a risk to the integrity of graphene. This study introduces a systematic strategy to protect graphene during the epitaxial growth of BaTiO3 (BTO) on SrTiO3 (STO) substrates coated with graphene. A novel approach is established based on an initial BTO growth phase using a laser source with a controlled aperture, allowing for the modulation of the growth rate to minimize damage to the underlying graphene layer. Perhaps for the first time, it is shown that a direct correlation exists between the grain size of graphene, the generation of defects subsequent to its exposure under the PLD plume, and the crystalline quality of the BTO film on top. The resulting BTO films exhibit gradual strain relaxation with the insertion of multiple graphene layers. Using bilayer graphene facilitates the easy exfoliation and transfer of the BTO film. These insights pave the way for the heterogeneous integration of diverse functional oxides, holding significant implications for commercializing perovskite oxides in flexible electronics. | en_US |
